RT info:eu-repo/semantics/article
T1 Towards a hierarchical approach for autotuning task-based libraries
A1 Cámara Moreno, Jesús
A1 Cuenca Muñoz, Javier
A1 Boratto, Murilo
K1 Programación de ordenadores
K1 Ingeniería de sistemas
K1 Tecnología de la información
K1 Ajuste automático jerárquico
K1 Computación heterogénea
K1 Programación basada en tareas
K1 1203 Ciencia de Los Ordenadores
K1 3304 Tecnología de Los Ordenadores
AB This work proposes a hierarchical approach to reduce the training time of task-based routines by reusing previously obtained autotuning information. This approach has been integrated into a working prototype of Chameleon, a dense linear algebra software whose tile-based routines are executed on the available computational resources by means of a runtime system. The results show that this approach provides a high degree of scalability to the entire self-optimization process, achieving a reduction in training time of up to 80% and an appropriate selection of values for the adjustable parameters.
PB Springer Nature
SN 0920-8542
YR 2026
FD 2026
LK https://uvadoc.uva.es/handle/10324/83751
UL https://uvadoc.uva.es/handle/10324/83751
LA eng
NO The Journal of Supercomputing, 2026, vol. 82, n. 5, artículo 271.
NO Producción Científica
DS UVaDOC
RD 24-mar-2026